ASMlib 技術(shù)應(yīng)用

R9_Alien

管理ASM磁盤組一定需要ASMlib這些支持庫么
http://www.itpub.net/viewthread.php?tid=930411&extra=&page=1
回答是不一定！
ASM using ASMLib and Raw Devices
http://www.oracle-base.com/articles/10g/ASMUsingASMLibAndRawDevices.php
ASMLib的用處：
Project: ASMLib Kernel Driver
http://oss.oracle.com/projects/oracleasm/
ASMLib allows an Oracle Database using ASM more efficient and capable access to the disk
groups it is using.
ASMlib不是必須的，但是安裝了可以增加ASM磁盤組管理的簡便性。
ASMLib 令使用 ASM 的 Oracle 數(shù)據(jù)庫更加高效，并能夠訪問它正在使用的磁盤組。
安裝 Oracle ASMLib
http://www.oracle.com/technology/global/cn/tech/linux/asmlib/install.html
檢查本機的內(nèi)核版本
[root@racdb1 ~]# uname -r
2.6.9-55.ELsmp
Migrating Raw Devices to ASMLib
http://www.oracle.com/technology/tech/linux/asmlib/raw_migration.html
1、安裝Oracle ASMLib
下載地址:http://www.oracle.com/technology/software/tech/linux/asmlib/rhel4.html
[root@racdb1 asm]# rpm -ivh oracleasm-support-2.0.3-1.i386.rpm
[root@racdb1 asm]# rpm -ivh oracleasm-2.6.9-55.0.12.EL-2.0.3-1.i686.rpm
[root@racdb1 asm]# rpm -ivh oracleasmlib-2.0.2-1.i386.rpm
2、配置ASMLib
[root@racdb1 asm]# /etc/init.d/oracleasm configure
Default user to own the driver interface []: oracle
Default group to own the driver interface []: dba
Start Oracle ASM library driver on boot (y/n) [n]: y
Fix permissions of Oracle ASM disks on boot (y/n) [y]: y
Writing Oracle ASM library driver configuration: [  OK  ]
Creating /dev/oracleasm mount point: [  OK  ]
Loading module "oracleasm": Unable to load module "oracleasm"
[FAILED]
最后出現(xiàn)錯誤,檢查
[root@racdb1 asm]# tail -100 /var/log/messages
[root@racdb1 asm]# rpm -qa | grep oracleasm
[root@racdb1 asm]# uname -rm
[root@racdb1 asm]# rpm -qa | grep kernel
發(fā)覺是安裝的版本不對，開始以為滿足kernel的大版本就可
[root@racdb1 asm]# rpm -e oracleasmlib-2.0.2-1
[root@racdb1 asm]# rpm -e oracleasm-2.6.9-55.0.12.EL-2.0.3-1
注：rpm若一定要刪除，可以使用：#rpm -e XXX --nodeps
下載安裝正確的版本：
[root@racdb1 asm]# rpm -Uvh oracleasm-2.6.9-55.EL-2.0.3-1.i686.rpm
[root@racdb1 asm]# rpm -Uvh oracleasmlib-2.0.2-1.i386.rpm
[root@racdb1 asm]# rpm -ivh oracleasm-2.6.9-55.ELsmp-2.0.3-1.i686.rpm
添加init文件使系統(tǒng)啟動時自動加載ASMLib
[root@racdb1 rc3.d]# ll *oracleasm
lrwxrwxrwx  1 root root 19 Mar  1 12:26 S29oracleasm -> ../init.d/oracleasm
#cd /etc/rc3.d
#ln -s ../init.d/oracleasm K01oracleasm
重新啟動系統(tǒng),確認ASMLib已經(jīng)可以自動加載
#lsmod |grep oracleasm
#dmesg |grep oracleasm
就我的理解，如果是單個磁盤，或容量比較小的，直接只用如 /dev/sda, /dev/sdb就可，
如果容量比較大，才需要分區(qū)（比如外置磁盤柜的lun），使用如/dev/sda1, /dev/sda2
修改磁盤所有者：
[root@racdb1 dev]# chown oracle /dev/sdd*
[root@racdb1 dev]# chmod 660 /dev/sdd*
[root@racdb1 dev]# ll /dev/sdd*
如果之前ASM已經(jīng)使用過該磁盤,可以清除掉head信息：
[root@racdb1 ~]# dd if=/dev/zero of=/dev/sdc bs=1M count=1024
[root@racdb1 ~]# dd if=/dev/zero of=/dev/sdd bs=1M count=2024
如果用fdisk之類工具對磁盤重新分區(qū)，讓系統(tǒng)重新掃描磁盤分區(qū)：
[root@racdb1 raw]# partprobe
創(chuàng)建ASM磁盤：
[root@racdb1 ~]# /etc/init.d/oracleasm createdisk VOL1 /dev/sdc1
Marking disk "/dev/sdc1" as an ASM disk: [  OK  ]
[root@racdb1 ~]# /etc/init.d/oracleasm createdisk VOL2 /dev/sdc2
Marking disk "/dev/sdc2" as an ASM disk: [  OK  ]
[root@racdb1 ~]# /etc/init.d/oracleasm createdisk VOL3 /dev/sdd1
Marking disk "/dev/sdd1" as an ASM disk: [  OK  ]
[root@racdb1 ~]# /etc/init.d/oracleasm createdisk VOL4 /dev/sdd2
Marking disk "/dev/sdd2" as an ASM disk: [  OK  ]
[root@racdb1 ~]# /etc/init.d/oracleasm listdisks
VOL1
VOL2
VOL3
VOL4
如果要刪除ASM磁盤通過以下命令
#/etc/init.d/oracleasm deletedisk VOL4
On all other RAC nodes, you just need to notify the system about the new ASM disks:
如果有多節(jié)點的rac
[root@racdb1 ~]# /etc/init.d/oracleasm scandisks
Scanning system for ASM disks: [  OK  ]
[root@racdb1 disks]# /etc/init.d/oracleasm status
Checking if ASM is loaded: [  OK  ]
Checking if /dev/oracleasm is mounted: [  OK  ]
注意mount在 /dev/oracleasm
安裝oracle, 打補丁
運行netca
運行dbca
選擇 asm
按提示運行, 安裝crs：
[root@racdb1 ~]# /u01/oracle/product/db_1/bin/localconfig add
創(chuàng)建磁盤組
參考：
http://www.oracle-base.com/articles/10g/AutomaticStorageManagement10g.php
磁盤組名稱
- 輸入磁盤組名稱，如 DATA1。
冗余
- High 鏡像數(shù)據(jù)兩次（三份數(shù)據(jù)）。
- Normal 鏡像數(shù)據(jù)一次（兩份數(shù)據(jù)）。
- External 不鏡像 ASM 中的數(shù)據(jù)。如果外部 RAID 陣列提供冗余，則通常使用此選項。
選擇成員磁盤
- 您必須手動指示 ASM 到什么地方查找 ASMLib 磁盤，因此，單擊 Change Disk Discovery Path
- 彈出一個窗口，提示輸入磁盤發(fā)現(xiàn)路徑。如下所示將磁盤發(fā)現(xiàn)路徑更改為"/dev/sd*". “ORCL:*”。
  在這里，我只有輸入"/dev/oracleasm/disks/VOL*"才能找到磁盤
- 單擊 OK 繼續(xù)。
您將返回到 Select Member Disks 窗口。您的所有 ASMLib 磁盤現(xiàn)在應(yīng)顯示在此窗口中。
單擊該行開頭的框，選擇加入磁盤組的磁盤。完成后單擊 OK。
管理ASM Disk Groups
登錄及啟動關(guān)閉:
$ ORACLE_SID=+ASM; export ORACLE_SID
$ sqlplus "/ as sysdba"
SQL> startup
SQL> shutdown immediate;
解決錯誤：
ORA-15032: not all alterations performed
ORA-15063: ASM discovered an insufficient number of disks for diskgroup "DATA1"
清除已經(jīng)移除物理磁盤的DISKGROUP
[oracle@racdb1 dbs]$ strings spfile+ASM.ora | more
SQL> alter system set asm_diskgroups=DGDATA scope=spfile;
SQL> alter system set asm_diskgroups=DGDATA scope=spfile sid='+ASM';
兩個很有用的檢查disk group信息的SQL，下面的設(shè)置會反復(fù)用到：
SQL> col state format a10
SQL> col name format a15
SQL> col failgroup format a20   
SQL> set line 200
SQL> select STATE,REDUNDANCY,TOTAL_MB,FREE_MB,NAME,FAILGROUP from v$asm_disk;
SQL> select GROUP_NUMBER,NAME,STATE,TYPE,TOTAL_MB,FREE_MB,UNBALANCED from v$asm_diskgroup;
-- 創(chuàng)建新的diskgroup
SQL> create diskgroup dgtest normal redundancy
  2  failgroup dgtest_0001 disk '/dev/oracleasm/disks/VOL5' name dgtest_0001
  3  failgroup dgtest_0002 disk '/dev/oracleasm/disks/VOL6' name dgtest_0002;
-- 刪除diskgroup
SQL> drop diskgroup dgtest including contents;
-- 為diskgroup增加disk
SQL> alter diskgroup dgdata add disk '/dev/oracleasm/disks/VOL5' name
dgdata_0004,'/dev/oracleasm/disks/VOL6' name dgdata_0005;
-- 從diskgroup刪除disk
SQL> alter diskgroup dgdata drop disk DGDATA_0004;
-- 取消刪除disk的命令，只能在上述命令沒執(zhí)行完成的時候有效
The UNDROP DISKS clause of the ALTER DISKGROUP statement allows pending disk drops to be
undone. It will not revert drops that have completed, or disk drops associated with the
dropping of a disk group.
ALTER DISKGROUP disk_group_1 UNDROP DISKS;
SQL> show parameter asm_power_limit
這個參數(shù) ASM_POWER_LIMIT 參數(shù)控制 ASM 后臺進程 ARBx 的數(shù)量。ARBx 進程用來進行 ASM 磁盤數(shù)據(jù)重新分布打散。ASM_POWER_LIMIT 取值 從 0 到 11(據(jù)說從 10gR2 開始可以設(shè)置為 0 ). 當新添加磁盤或者刪除磁盤后，ASM 會啟動 ARBx 進行 IO 分散操作，這是個非常消耗資源的動作，所以一定要選擇系統(tǒng)空閑的時候進行。
SQL> alter diskgroup dgdata rebalance power 5;
Disk groups can be rebalanced manually using the REBALANCE clause of the ALTER DISKGROUP statement. If the POWER clause is omitted the ASM_POWER_LIMIT parameter value is used. Rebalancing is only needed when the speed of the automatic rebalancing is not appropriate.
其它一些手動mount命令
ALTER DISKGROUP ALL DISMOUNT;
ALTER DISKGROUP ALL MOUNT;
ALTER DISKGROUP dgdata DISMOUNT;
ALTER DISKGROUP dgdata MOUNT;
數(shù)據(jù)文件的管理
Aliases allow you to reference ASM files using user-friendly names, rather than the fully
qualified ASM filenames.
-- Create an alias using the fully qualified filename.
ALTER DISKGROUP disk_group_1 ADD ALIAS '+disk_group_1/my_dir/my_file.dbf'
  FOR '+disk_group_1/mydb/datafile/my_ts.342.3';
-- Create an alias using the numeric form filename.
ALTER DISKGROUP disk_group_1 ADD ALIAS '+disk_group_1/my_dir/my_file.dbf'
  FOR '+disk_group_1.342.3';
-- Rename an alias.
ALTER DISKGROUP disk_group_1 RENAME ALIAS '+disk_group_1/my_dir/my_file.dbf'
  TO '+disk_group_1/my_dir/my_file2.dbf';
-- Delete an alias.
ALTER DISKGROUP disk_group_1 DELETE ALIAS '+disk_group_1/my_dir/my_file.dbf';
-- Drop file using an alias.
ALTER DISKGROUP disk_group_1 DROP FILE '+disk_group_1/my_dir/my_file.dbf';
-- Drop file using a numeric form filename.
ALTER DISKGROUP disk_group_1 DROP FILE '+disk_group_1.342.3';
-- Drop file using a fully qualified filename.
ALTER DISKGROUP disk_group_1 DROP FILE '+disk_group_1/mydb/datafile/my_ts.342.3';
-- create datafile
SQL> create tablespace users2 datafile '+TESTDB_DATA1' size 100m;

本文來自ChinaUnix博客，如果查看原文請點：http://blog.chinaunix.net/u/12437/showart_1953130.html